[PATCH] mmap zero-length hugetlb file with PROT_NONE to protect a hugetlb virtual...
commitb6174df5eec9cdfd598c03d6d0807e344e109213
authorZhang, Yanmin <yanmin.zhang@intel.com>
Mon, 10 Jul 2006 11:44:49 +0000 (10 04:44 -0700)
committerLinus Torvalds <torvalds@g5.osdl.org>
Mon, 10 Jul 2006 20:24:21 +0000 (10 13:24 -0700)
treed61c8627138a8feee31de8320e337251d567fca9
parentd6b7d3b62069be60d5b13358bac8670dacdd7a81
[PATCH] mmap zero-length hugetlb file with PROT_NONE to protect a hugetlb virtual area

Sometimes, applications need below call to be successful although
"/mnt/hugepages/file1" doesn't exist.

fd = open("/mnt/hugepages/file1", O_CREAT|O_RDWR, 0755);
*addr = mmap(NULL, 0x1024*1024*256, PROT_NONE, 0, fd, 0);

As for regular pages (or files), above call does work, but as for huge
pages, above call would fail because hugetlbfs_file_mmap would fail if
(!(vma->vm_flags & VM_WRITE) && len > inode->i_size).

This capability on huge page is useful on ia64 when the process wants to
protect one area on region 4, so other threads couldn't read/write this
area.  A famous JVM (Java Virtual Machine) implementation on IA64 needs the
capability.

Signed-off-by: Zhang Yanmin <yanmin.zhang@intel.com>
Cc: David Gibson <david@gibson.dropbear.id.au>
Cc: Hugh Dickins <hugh@veritas.com>
[ Expand-on-mmap semantics again... this time matching normal fs's. wli ]
Acked-by: William Lee Irwin III <wli@holomorphy.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
fs/hugetlbfs/inode.c